^{page 274 note 2} See my The Somali Lineage System and the Total Genealogy, Hargeisa (British Somaliland), and Crown Agents for Colonies, London, 1957Google Scholar.

^{page 274 note 3} In the orthography adopted in this paper, long vowels are indicated by doubling, as, e.g. aa, oo, &c. The Arabic and Cushitic aspirate h is represented by h; the voiced pharyngal fricative by ‘; the glottal stop by’; and the Cushitic post-alveolar plosive

^{page 274 note 4} The Ishaaq are included in this estimate.

^{page 275 note 1} The Ḍulbahante clan comprises some 15 primary lineage-groups and some 50 Dia-paying groups. Inàfew cases, small primary lineage-groups which are not highly segmented act as Dia-paying groups. This is true of, e.g. the Hayaag, Yaḥya, Ḥaamud Ugaas, &c.

^{page 275 note 2} From ‘Igaal Faarah's eight sons stem the lineages shown. They are divided into two uterine groups, the Bah Hawiye and Bah Ugaad, and the latter are joined byàvery small group, the Reer Haad Faaraḥ, descended from ‘Igaal's brother, Haad.

^{page 276 note 1} See Africa, July 1958, pp. 244–61.

^{page 276 note 2} At all levels of segmentation there isàgeneral correlation between the size and strength ofàlineage-progroup and its genealogical span.

^{page 276 note 3} This isàSomali clan, known officially as ‘Arap’ to distinguish it from ‘Arab (an Arabian), which expatriate officials often find difficulty in pro-nouncing.

^{page 277 note 1} The Indian loan-word javaabdaar which was current during the administration of the coast by the Indian Government (1885–98) is still sometimes used today. Jawaab means ananswer, and Jarvaabdaar someone who answers for, or is responsible for.

^{page 277 note 2} The Arabic expression is rarely used in Northern Somaliland, but the Government has adopted it in naming the Dia-paying groups.

^{page 277 note 3} Evans-Pritchard, E. E., The Nuer, London, Oxford University Press, 1940Google Scholar.

^{page 278 note 1} Fortes, M., The Dynamics of Clanship among the Tallensi, London, Oxford University Press, 1945Google Scholar.

^{page 278 note 2} Cf. Arabic, q-b-y-1-t; in Somali qabiila also occurs.

^{page 278 note 3} In distinction to the other terms, all meaning ‘group’ inàgeneral sense and particularly agnatic group, jilib means literallyàjoint or part, as, e.g.,group,àknee.

^{page 278 note 4} See the genealogy above, p. 275.

^{page 279 note 1} The word reer is also generally applied toàhamlet, referring particularly to its composition asàgroup of people of the same lineage rather than simplyàphysical structure of huts arranged withinàcommon fence.

^{page 279 note 2} Inàsample of 58 hamlets containing 182 nuclear families the median hamlet contained 2.6 huts and nuclear families. The commonest size is between two and four nuclear families, although in time of war larger hamlets often occur.

^{page 280 note 1} Immisaawow yey kala tirsinayaan, from awow, generation or grandfather, and by extension, ancestor.

^{page 280 note 2} From ab or aabbe, father, tiri, to count, and sii, to give.

^{page 280 note 3} The expression Qolomaad tahay is also used, from qolo, see above, p. 278.

^{page 281 note 1} Thus the expression ḥayn iyo ḥiniin, literally the cloth worn by women round their limbs and the testicle, means ‘relatives through women and agnates’.

^{page 281 note 2} These words are from ḥig, to be near. There isànoun ḥig which meansàspecies of aloe(Sansevieria abyssinica), and also the second milking of livestock. I was not able to discover any connexion between these. Ḥig occurs in the expression ḥigsiisan, foràwife's sister or other close agnate married byàman after his wife's death. The expression means ’the one given next’. The sororate, is not now, however, extensively practised.

^{page 282 note 1} The agnatic framework of nomadic Somali society, if not fully understood in its segmentary implications, has long been appreciated. One of the earliest writers to stress the lineage character was Ciamarra, G., Camera dei Deputati Doce XXXVIII, Relazione sulla Somalia Italiana, allegato Rome, B., 1911Google Scholar . Much has similarly been written on Somali ‘custom’. The most notable contributions are Cerulli, E., ‘II diritto consuetudinario della Somalia Italiana settentrionale’, Bollettino della Societa Africana d'ltalia, anno xxxviii, Naples, 1919Google Scholar ; Colucci, M., Principi di diritto consuetu dinario della Somalia Italiana meridionale, Florence, 1924Google Scholar ; Wright, A. C. A., ‘The Interaction of Various Systems of Law and Custom in British Somaliland and their Relation with Social Life’, Journal of the East African Natural History Society, xvii. 1–2, 1943, pp. 66–102Google Scholar . The first indica-tion, however, as far as I am aware, in the literature of the true nature of Somali heer isàpaper by Drysdale, J. G. S. of the Protectorate Administration, (‘Some Aspects of Somali rural society today’, Somaliland Journal, Hargeisa [British Somaliland], i. 2,1955Google Scholar .) Most writers, and there have been many, have failed to appreciate the fundamentally contractual nature of heer and its interaction with agnation.

^{page 282 note 2} ‘There is an agreement between us’, Heer baainnagu deheeya, isàcommon way of stating this position.

^{page 282 note 3} The Arabic loan-word qaynuun (Ar. q-a-n-w-n)is used generally for custom or law, especially civillaw, as applied inàGovernment court.

^{page 282 note 4} The most often used work is , Al-Nawawii'sMinhaaj at-Taalibiin; for other Shariah sources see Bulletin of the School of Oriental and African Studies, 1956, xviii/i, p. 150Google Scholar.

^{page 282 note 5} Although the rates most generally current today in Northern Somaliland are for the homicide ofàman 100 camels and foràwoman 50 camels (ortheir equivalent in other currency), in parts of French Somaliland and Northern Somalia local agreements have established lower values of blood-wealth. An additional amount may be added to the basic rate where an important and generally well-known and respected man has been killed, or whereàmurder disturbsàlong-standing pact of friendship, bringing dishonour to the offending group.

^{page 283 note 1} Except between kinsmen of the nuclear or extended family.

^{page 283 note 2} Those appointed as arbitrators are known as heerbeeg, lit. those expert in custom (heer); gar means award, decision, justice, understanding, &c. The Arabic hukun is generally used rather of sanctioned judgement as administered byàqaaddi's or other government court.

^{page 283 note 3} Thus, for example, when the Reer Kooshin and the Reer Mahamuud 'Ali (two Dulbahante lineages)renounced their treaty, the seceding Reer Mahamuud 'Ali paid the Reer Kooshinàsum of over £70 sterling.

^{page 283 note 4} Applied under the terms of the Somaliland Principal Order in Council, 1929.

^{page 284 note 1} This procedure is the local version of the well-known Shafi'i qasaama. See Anderson, J. N. D., Islamic Law in Africa, H. M.S. O., 1954, p. 372Google Scholar.

^{page 284 note 2} The injuries have to be examined byàGovern- contriment doctor who sends the claimant toàqaaddi with note of their nature. The latter then assess damages,referring to Shafi'i textbooks.

^{page 284 note 3} About 1920, when the normal Somali practice of requiring an apprehended thief to return the article stolen (sometimes twice over) proved ineffective, bloodthe Government was asked to apply the Indian Penal Code.

^{page 284 note 4} Lit. the greater blood-wealth. In the north-west the expression qubarte is current.

^{page 284 note 5} Women are not fully members of the Dia-paying groups of their husbands. The responsibility for their safety and for injuries committed by them lies partly with their own agnates. To some extent the practice varies from clan-family to clan-family. Among the Ishaaq of the Protectorate generally, ifàmarried woman is killed, whether or not she has borne children, her agnatic kin and her husband's agnates divide her blood-wealth equally between them. Similarly, whereàmarried woman kills someone, her husband's agnates and her own agnates contribute equally to the blood-wealth due. The husband is, however, with his kin, solely responsible for small delicts committed by his wife. Among the Daarood and ‘Iise, on the other hand,àmarried woman's agnates are solely responsible when she kills someone, and reciprocally receive her blood-wealth if she is killed. But ifàmurdered woman leaves behind children, these are entitled to share in her blood-wealth with her own agnates. In small amounts of compensation for whichàmarried woman is responsible, payment is made out of her personal dower (meher), or paid by her children if they are adult, or in some cases paid by the husband, Generally, among both the Isḥaaq and Daarood, I believe, ifàwife kills her husband his blood-wealth is paid by her agnatic kin to his agnates and the children of the deceased usually receive the amount of the jiffo. Whenàman kills his wife, full blood-wealth is paid by his agnates to his wife's kin. Cases of this kind are rare, although I have come across examples.

^{page 285 note 1} Meel means literally ‘place’.

^{page 286 note 1} see my ‘The Gadabuursi Somali Script’, Bulletin of the School of Oriental and African Studies, 1958, xxi/i, pp. 135–40Google Scholar.

^{page 287 note 1} A remark inàpaper by Cruttenden (‘Memoir of the Western or Edoor tribes, in habiting the Somalicoast of North-East Africa’, Journal of the Royal Geographical Society, xix, 1849, PP. 49–76Google Scholar) suggests that peace was sometimes made atàprominent saint's shrine and treaties of friendship lodged with the custodians of the shrine. But, on the whole, the evidence is not very satisfactory.

^{page 287 note 2} This is known as jaa'ifo (as distinct from jiffo)and is any non-fatal but fairly serious wound for which compensation is of the order of thirty-three- and-a-third camels, but may be more.

^{page 287 note 3} When two men attack another of the same group this is known as hiil or tuuto and insult compensation is regularly payable. This isàcommon provision in Dia-paying treaties.

^{page 287 note 4} This refers to the Somali practice of widow inheritance (dumaal), where the children belong to the new husband and not to the deceased brother or close agnate (it is not the true levirate).àreduced bride-wealth is normally paid.

^{page 288 note 1} 1 See above, p. 285.

^{page 288 note 2} For information on these treaties I am much indebted to the courtesy of Mr. H.Y.W. S.Dickson, District Commissioner at Las Anod, in allowing me access to the District Office files.

^{page 288 note 3} Here, as in other places, I have paraphrased the actual wording in the interests of clarity in consultation with informants.

^{page 288 note 4} In north-western Somali, the pronunciation is badigpoyo. Amongst the eastern Dulbahante the Cushitic post-alveolar d has been modified to r, see Moreno, M. M., II Somalo della Somalia, Rome, 1955, p. 11Google Scholar . The word literally means ‘cutting off the entire property’.

^{page 289 note 1} I use this term to mean agnation in general, at every level of lineage (and clan) affiliation.

^{page 289 note 2} Marriage is mainly virilocal, and most marriages begin virilocally.

^{page 289 note 3} The whole question of adoption and assimilated status is complex and I can here indicate the position only inàvery generalway. The subject will be examined more fully inàforthcoming publication.

^{page 289 note 4} Collectively called sab.

^{page 290 note 1} 1 There areàfew exceptions where sab groups have achieved recognition by the Administrations as independent Dia-paying units. There is todayàgeneral movement in this direction, the sab seeking to emancipate themselves from their traditional bonds to their patrons.

^{page 290 note 2} Buur ahaw ama buur ku tirso.

^{page 290 note 3} Literally, ‘common or general heer’; Irleer do'nimo means literally local, or specific, heer.

^{page 291 note 1} Professor Gluckman has kindly drawn my attention to Barton's, R. F. accounts of the Ifugao and Kalingas (Ifugao Law, University of California PublitionGoogle Scholar cations in American Archaeology and Ethnology, vol.xv, pp. 1–127, 1919; The Kalingas, Chicago, 1949)Google Scholar . But the treaties or ‘pacts’ of the Kalingas establish relations between existing political groups: they do not, as among the Somali, create political units; nor do they determine the political and jural status of the individual. More closely analogous to the Somali system are the ‘asaba blood-compensation groups of the nomads of the Spanish Sahara. For here, as among the Somali, Dia-paying groups are established by verbal treaties among agnates. Yet'asaba does not appear to have the pervasive character Conof Somali heer. See Baroja, Julio Caro, Estudios Saharianos, Madrid, 1955, pp. 17–22Google Scholar.

^{page 292 note 1} Gluckman, M., The Judicial Process among the Barotse of Northern Rhodesia, ManchesterUniversity Press, 1955, p. 28Google Scholar.

^{page 292 note 2} Ancient Law (World's Classics edition), 1954, 141.

^{page 292 note 3} Edited and translated by R. Basset, Paris, 1897–1909.

^{page 292 note 4} See my ‘Modern Political Movements in Somaliland’, Africa, July and October 1958.

^{page 292 note 5} Foràconvenient historical discussion of contractual theories, see Gough, J. W., The Social Contract, Oxford, 1957Google Scholar.